ASE alumnus Melanie Kühnast recently co-authored an article on this topic, published on the Dutch economics platform ESB. The article is based on Kühnast’s Master’s thesis at the University of Amsterdam. She graduated in 2024 from the Amsterdam School of Economics (ASE) with an MSc in Economics, specialising in Public Policy.

What was the NOW scheme?

The NOW (Temporary Emergency Bridging Measure for Sustained Employment) helped businesses pay their employees’ wages if their revenue dropped by at least 20% due to COVID-19. The scheme was rolled out in phases – from NOW1 to NOW6 – each with slightly different conditions. The first phase, NOW1, ran from March to June 2020 and was the largest: around 140,000 companies received a total of €5.2 billion in support.

Companies applied for the subsidy based on their expected revenue loss. Later, they had to report their actual revenue loss. If the real loss was lower than expected, they had to repay part of the subsidy. If it was higher, they received an additional payment.

Opportunities for strategic behaviour

This setup gave companies room to tweak their numbers. If they reported a revenue loss of less than 20%, they had to pay everything back. And the greater the reported loss, the more support they received. So, there was a clear incentive to make the loss seem bigger than it really was.

Kühnast investigated whether companies actually did this. She combined data from the UWV (the Dutch Employee Insurance Agency) with VAT figures from Statistics Netherlands (CBS). This allowed her to compare the self-reported revenue loss with the actual turnover filed with the tax authorities.

What were the findings?

Almost 40% of the companies that received NOW1 support reported a higher revenue loss than what their VAT filings showed. Many of them reported losses just above the 20% threshold – exactly enough to qualify for the subsidy. That’s a strong sign of strategic behaviour. If these findings are extrapolated to all NOW1 recipients, it would amount to roughly €740 million in excess support.

Is that fraud?

Not necessarily. Revenue can be difficult to measure, especially during a crisis. Some discrepancies may be explainable. But the pattern – many reported losses just above the threshold – suggests that companies knew how to work the system to their advantage. Some likely stayed within the rules, others may have crossed the line. 
There have already been confirmed cases of misuse. Temp agencies were investigated, and business owners prosecuted. In 2024, the Dutch Labour Inspectorate even arrested several individuals suspected of subsidy fraud.

Why does this matter?

The NOW scheme achieved a lot: it prevented mass layoffs and kept businesses afloat. But it’s also important to learn from what didn’t go so well. ‘Policy evaluation is essential,’ Kühnast says. ‘People respond to incentives, and that affects not only how effective support measures are, but also their cost to society and how public resources are distributed.’

Her findings show how crucial it is to design policies in a way that discourages abuse – even when speed is of the essence.